Large Scale Antenna Systems (Massive MIMO) Capacity( bτs) = NBlog 2 1 + S N + I Contiguous available bandwidth Additional channels due to huge number of antennas Optimized Signal to Noise+Interference ratio due to adaptive beam forming Tongyu Communication Inc. Dr. Doudou Samb, Base Station Antenna R&D PL Technical Director, 4.5G/5G Lead
Problems or Challenges: Current Antenna Systems? High-rise building coverage: Limited directive antennas (in azimuth/elevation plan) resulting on limitation in terms of high order sectorization. Capacity lift @Macro Site and Uplink Coverage& Capacity Limited: For a given allocated time-frequency, there is still challenge during multiplexing of different users due to small number of available antennas being able to direct azimuth narrow beam at desired direction while nulling interferers of intraand inter-cell efficiently. Besides, business expansion along with difficulty in acquiring new site where UL:DL is 1:3 High In-Building Capacity growth: Even in claimed SU-MIMO, resources are not exploited fully due to limited size of user devices. Besides, Higher cost for in-building system, with poor WLAN performance. Efficient management covering as much frequency bands as possible Low latency with significant capacity out of a given flexible bandwidth Coverage not big issue but capacity is! Low tower load with acceptable dimensions for multiple service applications System upgrade needed, but worrying about impact on live networks Solution:3D-MIMO via Large Scale Antenna Systems
Beam forming On the Go: 3D-MIMO One antenna to realize coverage in low and tall building, high beam forming to realize penetration resistance Accurate and flexible tri-angle beam forming to support more users MU -BF - Increased Spectrum efficiency by Smart collocated or conformal antenna arrays along with Vertical beam adjustment. - Key technology driving 4.5G/5G recently. - Standardization should be promote with effort, prototype along with network deployment pilot. - In the Long-term, beam forming in higher frequency and hardware progress can be considered
Massive MIMO Techniques Multi-beam antenna array( multi-input, simultaneous multi-beam antenna array) 1. Passive special feed network:butler matrix,rotman lens 2. Stacked-beam (electrical-large antennas (eg, reflector or dielectric lens) with multi-source stimulated,) 3. Digital phased array Passive phased array Active phased array Digital phased array ANT ANT ANT ANT ANT ANT Phase shifter Phase shifter Phase shifter Analog signal Low power signal Digital signal
Potential Scheme Principle and solution: - With 8*8 array,all together 64 antenna units,output 64 RF ports and 1 calibration port; - Antenna array consists of 4 parts of sub-array module including antenna units, feeding network and 16 to 1 calibration network; - Each sub-array module consists of 16 antenna units,(4 lines 4 rows)and 1 set of 16 to 1 calibration network; every two antenna units through one to two splitter, combine to one RF port, and output 16 RF ports(including eight +45 polarization ports and eight -45 polarization ports ) and 1 calibration sub-port; - 4 sub-array module output four 16 to 1 calibration ports, and realize 4 to 1 function on the back of antenna,so that to realize whole antenna function of 64 to 1 calibration Antenna feeding network
Coupling Characterization Model Out1... Let s denote a0 the incident wave to the antenna input (corresponds also to the feeding network input) and r 0 the corresponding reflected wave. We model also ai as the incident wave to the i th output of the feeding network, r i being the i th reflected wave from the i th antenna element to the i th output of the feeding network. Thus, by denoting S as the scattering matrix, the parameter relation for the feeding network can be derived as: th i OutN S r0 a0 S r a i i S S 00 0i i0 S S ii Where S 00 is the reflection coefficient of the feeding network, S0i Si0is a 1xN sub-matrix characterizing the power transfer vector to the feeding network outputs and S ii a NxN submatrix characterizing the coupling relation of the feeding network outputs. The array structure is designed and optimized using HFSS. In this experiment an 11-elements is considered as can be seen from fig.2. And the corresponding array S-parameters Sarr and each element s pattern Pi can be obtained.
Simulation and Measurement Results Analysis etilt EBW SLL(0-30 ) SLL1(First) SLL2(Max) TEST MC-F TEST MC-F TEST MC-F TEST MC-F TEST MC-F 1700 0.85 1 7.18 7.34 21.12 17.42 21.12 17.42 21.12 17.42 1800 0.85 0 6.88 7.01 17.32 21.04 25.53 22.54 17.32 21.04 1900 0.85 1 6.39 6.63 18.79 20.83 24.65 21.86 18.79 20.83 2000 1.27 0 6.22 6.32 18.38 24.53 29.54 26.25 18.38 22.89 2100 0.85 1 6.16 5.85 21.69 15.96 29.07 20.48 19.92 15.96 2200 0.85 0 5.91 5.75 22.91 18.44 22.91 22.25 19.36 18.44 2300 1.06 0 5.43 5.49 18.61 19.78 23.9 21.22 17.8 19.78 2400 0.85 0 5.56 5.37 15.71 17.36 15.71 17.36 15.71 17.36 2500 0.43 0 5.28 5.11 19.38 18.32 19.38 18.32 17.15 17.5 2600 1.06 0 5.01 4.82 19.46 18.53 23.22 20.69 12.43 12.24 2700 0.85 0 4.58 4.61 14.26 16.58 14.26 16.58 10.58 10.36 MAX 1.27 1.00 7.18 7.34 22.91 24.53 29.54 26.25 21.12 22.89 MIN 0.43 0.00 4.58 4.61 14.26 15.96 14.26 16.58 10.58 10.36 AVG 0.89 0.27 5.87 5.85 18.88 18.98 22.66 20.45 17.14 17.62
Potential Scheme Vertical-plane Feeding-Networks Horizontal-plane Feeding-Networks Array layout Port Cable Cable Vertical-palne Cable Horizontal-palne Matrix Switch Outline
Platform Schematic (Scheme 2) Matrix Switch Outline Multi beam network port, ie, beam port 4 for 1 Microwave switch RF cable RF cable RF cable RF cable RR U Loa d RRU RF Control cable line RRU Control Port
Simulation Results H-Plane 7 beams V-Plane 8 beams 1dB 1dB